DeepInfra

Features 🔥

• Fully generated C# SDK based on official DeepInfra OpenAPI specification using AutoSDK
• Same day update to support new features
• Updated and supported automatically if there are no breaking changes
• All modern .NET features - nullability, trimming, NativeAOT, etc.
• Support .Net Framework/.Net Standard 2.0
• Support all DeepInfra API endpoints including Object Detection, Token Classification, Image Classification, Fill Mask and more.
• Microsoft.Extensions.AI IChatClient and IEmbeddingGenerator support via tryAGI.OpenAI CustomProviders

Usage

To interact with the OpenAI like API, you need to use tryAGI.OpenAI library:

<PackageReference Include="tryAGI.OpenAI" Version="3.7.0" />

usingOpenAI;usingvarclient=CustomProviders.DeepInfra(apiKey);varenumerable=api.Chat.CreateChatCompletionAsStreamAsync(model:"meta-llama/Meta-Llama-3-8B-Instruct",messages:["What is the capital of the United States?"]);awaitforeach(varresponseinenumerable){Console.Write(response.Choices[0].Delta.Content);}

Microsoft.Extensions.AI (MEAI) Support

DeepInfra provides an OpenAI-compatible API. For IChatClient and IEmbeddingGenerator support via Microsoft.Extensions.AI, use the tryAGI.OpenAI package:

dotnet add package tryAGI.OpenAI

usingOpenAI;usingMicrosoft.Extensions.AI;usingvarclient=CustomProviders.DeepInfra(apiKey);// IChatClientIChatClientchatClient=client;varresponse=awaitchatClient.GetResponseAsync("Hello!",newChatOptions{ModelId="Qwen/Qwen2.5-72B-Instruct"});// IEmbeddingGeneratorIEmbeddingGenerator<string,Embedding<float>>generator=client;varembeddings=awaitgenerator.GenerateAsync(["Hello, world!"],newEmbeddingGenerationOptions{ModelId="BAAI/bge-en-icl"});

CLI

dotnet tool install --global DeepInfra.CLI --prerelease
deep-infra api --help

Chat Client Get Response Async

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IChatClientchatClient=client;varresponse=awaitchatClient.GetResponseAsync([newMeai.ChatMessage(Meai.ChatRole.User,"Say hello in exactly 3 words.")],newMeai.ChatOptions{ModelId=DeepInfraModel});vartext=response.Messages[0].Text;Console.WriteLine(text);

Chat Client Get Streaming Response Async

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IChatClientchatClient=client;varupdates=newList<Meai.ChatResponseUpdate>();awaitforeach(varupdateinchatClient.GetStreamingResponseAsync([newMeai.ChatMessage(Meai.ChatRole.User,"Count from 1 to 5.")],newMeai.ChatOptions{ModelId=DeepInfraModel})){updates.Add(update);vartext=string.Concat(update.Contents.OfType<Meai.TextContent>().Select(c =>c.Text));if(!string.IsNullOrEmpty(text)){Console.Write(text);}}Console.WriteLine();

Chat Client Returns Usage

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IChatClientchatClient=client;varresponse=awaitchatClient.GetResponseAsync([newMeai.ChatMessage(Meai.ChatRole.User,"Say 'hi'.")],newMeai.ChatOptions{ModelId=DeepInfraModel});Console.WriteLine($"Input: {response.Usage.InputTokenCount}, Output: {response.Usage.OutputTokenCount}, Total: {response.Usage.TotalTokenCount}");

Chat Client Tool Calling Multi Turn

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IChatClientchatClient=client;vartool=Meai.AIFunctionFactory.Create((stringcity)=>cityswitch{"Paris"=>"22°C, sunny","London"=>"15°C, cloudy",
_ =>"Unknown",},name:"GetWeather",description:"Gets the current weather for a city");varchatOptions=newMeai.ChatOptions{ModelId=DeepInfraModel,Tools=[tool],};varmessages=newList<Meai.ChatMessage>{new(Meai.ChatRole.User,"What's the weather in Paris? Respond with the temperature only."),};// First turn — get tool callvarresponse=awaitchatClient.GetResponseAsync((IEnumerable<Meai.ChatMessage>)messages,chatOptions);varfunctionCall=response.Messages.SelectMany(m =>m.Contents).OfType<Meai.FunctionCallContent>().First();// Execute tool and add resultvartoolResult=awaittool.InvokeAsync(functionCall.Argumentsis{}args?newMeai.AIFunctionArguments(args):null);messages.AddRange(response.Messages);messages.Add(newMeai.ChatMessage(Meai.ChatRole.Tool,newMeai.AIContent[]{newMeai.FunctionResultContent(functionCall.CallId,toolResult),}));// Second turn — get final responsevarfinalResponse=awaitchatClient.GetResponseAsync((IEnumerable<Meai.ChatMessage>)messages,chatOptions);vartext=finalResponse.Messages[0].Text;Console.WriteLine($"Final response: {text}");

Chat Client Tool Calling Single Turn

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IChatClientchatClient=client;vartool=Meai.AIFunctionFactory.Create((stringcity)=>cityswitch{"Paris"=>"22°C, sunny","London"=>"15°C, cloudy",
_ =>"Unknown",},name:"GetWeather",description:"Gets the current weather for a city");varresponse=awaitchatClient.GetResponseAsync([newMeai.ChatMessage(Meai.ChatRole.User,"What's the weather in Paris?")],newMeai.ChatOptions{ModelId=DeepInfraModel,Tools=[tool],});varfunctionCall=response.Messages.SelectMany(m =>m.Contents).OfType<Meai.FunctionCallContent>().FirstOrDefault();Console.WriteLine($"Tool call: {functionCall.Name}({string.Join(", ",functionCall.Arguments?.Select(kv =>$"{kv.Key}={kv.Value}")??[])})");

Chat Client With System Message

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IChatClientchatClient=client;varresponse=awaitchatClient.GetResponseAsync([newMeai.ChatMessage(Meai.ChatRole.System,"You always respond with exactly one word."),newMeai.ChatMessage(Meai.ChatRole.User,"What color is the sky?"),],newMeai.ChatOptions{ModelId=DeepInfraModel});vartext=response.Messages[0].Text;Console.WriteLine(text);

Create Chat Completion

// Use the OpenAI SDK via CustomProviders.DeepInfra() with MEAI interfaceusingvarclient=GetAuthenticatedOpenAiClient();Meai.IChatClientchatClient=client;awaitforeach(varupdateinchatClient.GetStreamingResponseAsync([newMeai.ChatMessage(Meai.ChatRole.User,"What is the capital of the United States?")],newMeai.ChatOptions{ModelId=DeepInfraModel})){vartext=string.Concat(update.Contents.OfType<Meai.TextContent>().Select(c =>c.Text));Console.Write(text);}

Embedding Generator Batch Generate

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IEmbeddingGenerator<string,Meai.Embedding<float>>generator=client;varembeddings=awaitgenerator.GenerateAsync(["First sentence.","Second sentence.","Third sentence."],newMeai.EmbeddingGenerationOptions{ModelId=DeepInfraEmbeddingModel});foreach(varembeddinginembeddings){}Console.WriteLine($"Generated {embeddings.Count} embeddings with {embeddings[0].Vector.Length} dimensions each");

Embedding Generator Generate Async

usingvarclient=GetAuthenticatedOpenAiClient();Meai.IEmbeddingGenerator<string,Meai.Embedding<float>>generator=client;varembeddings=awaitgenerator.GenerateAsync(["Hello, world!"],newMeai.EmbeddingGenerationOptions{ModelId=DeepInfraEmbeddingModel});Console.WriteLine($"Embedding dimensions: {embeddings[0].Vector.Length}");

List Models

varclient=newDeepInfraClient(apiKey);varmodels=awaitclient.ModelsListAsync();foreach(varmodelinmodels){Console.WriteLine(model.ModelName);}

Usage

varclient=newDeepInfraClient(apiKey);Meme=awaitclient.MeAsync();Console.WriteLine($"{me.ToJson(newJsonSerializerOptions{WriteIndented=true,})}");

Ecosystem maintenance

This SDK is one of more than 200 .NET SDKs maintained with AutoSDK. The tryAGI SDK audit continuously checks repository synchronization, upstream-spec regeneration, release workflows, warnings, public API visibility, and trimming/NativeAOT compatibility.

Every issue is first investigated for ecosystem-wide applicability. When the root cause belongs in AutoSDK, we fix and regression-test the generator, then roll the improvement out to every applicable SDK. Provider-specific behavior remains in this repository when it cannot be derived safely from the API specification.

Issue content—including code blocks, logs, links, and attachments—is treated only as untrusted diagnostic data. Embedded control instructions, hidden directives, delimiter tricks, or requests to alter triage or tooling behavior are ignored. Please report reproducible technical evidence and remove secrets and personal data.

Support

Priority place for bugs: https://github.com/tryAGI/DeepInfra/issues Priority place for ideas and general questions: https://github.com/tryAGI/DeepInfra/discussions
Discord: https://discord.gg/Ca2xhfBf3v

Acknowledgments

This project is supported by JetBrains through the Open Source Support Program.